- Conference Organizers
- Registration Information
- Purchase the Box Set
- Why Attend LISA '13?
- Convince Your Boss
- Co-located Events
- Students and Grants
- Sponsorship and Exhibition
- Call for Participation
- For Participants
- Help Promote!
- Past Conferences
You are here
SRE Classroom: Non-Abstract Large System Design for Sysadmins
Thurgood Marshall North Ballroom
Truly large-scale systems are still rare; in a world of outsourcing and cloud computing, it's harder for system administrators to get the opportunity to design large systems. It's even harder to get the design right. Most organizations don't have the in-house expertise to build a large system and thus outsource the detailed design to external contractors. If your organization doesn't have the expertise to design a large system, it's unlikely that it has the expertise to confirm a proposal is fit for purpose and cost-effective.
While anyone can wave their hands convincingly and come up with a rough outline of a large distributed system, those who also can fill in the detail are highly prized. This class will teach you how to design software systems like Imgur and Twitter, then estimate the hardware needed to ensure you can deliver to an SLA. You will learn how requirements like queries-per-second, multi-site reliability, and data-security impact the cost of implementation.
This will not cover concurrent programming and the software-engineering side of distributed systems.
System administrators, SREs, and DevOps who have some familiarity with distributed systems, server hardware, and systems programming, especially those who would like to work with, procure, or build large distributed systems.
The ability to design large distributed systems, to evaluate design proposals, and to explain such designs to third parties.
- Design patterns for large distributed systems
- Monitoring large-scale systems
- Large-scale design workshop and presentations
- Non-abstract design; taking a design and producing a "bill of materials"
- Designing for failure; how to work around rack, networking, and datacenter failures